> ## Documentation Index
> Fetch the complete documentation index at: https://docs.secapi.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Search SEC Filing Content

> Choose full-text, semantic, or section search for SEC filing content, then carry accession numbers, source URLs, and citation spans into the next step.

SEC API has three different search jobs. Choose the one that matches how you will use the result:

* **Full-text search** finds both filing records and matching sections for an exact query.
* **Semantic search** retrieves conceptually related sections when the wording may differ.
* **Section search** searches extracted filing sections and returns citation-ready passages.

Search is retrieval, not a conclusion. Read the returned passage, preserve its accession number and source URL, and let an investor or downstream policy decide what the disclosure means.

## Full-text search: exact terms across filings and sections

```bash theme={null}
curl "https://api.secapi.ai/v1/search/fulltext?q=revenue%20recognition%20policy&ticker=AAPL&form=10-K&limit=5" \
  -H "x-api-key: $SECAPI_API_KEY"
```

The response contains both `filings` and `sections` lists. Use this when you want exact-term matches and the filing record around them.

## Semantic search: related disclosures, not only exact words

```bash theme={null}
curl "https://api.secapi.ai/v1/search/semantic?q=supply%20chain%20disruption%20risk&ticker=MSFT&form=10-K&mode=hybrid&limit=5" \
  -H "x-api-key: $SECAPI_API_KEY"
```

`mode` can be `keyword`, `semantic`, or `hybrid`; the default is `hybrid`. The response has a `sections.data` list. In the default view, each result includes retrieval metadata. `view=agent` keeps the citation fields while omitting score and retrieval-mode metadata where the endpoint documents that projection.

## Section search: a cited passage for a known research question

```bash theme={null}
curl "https://api.secapi.ai/v1/sections/search?q=going%20concern&ticker=AAPL&form=10-K&limit=5&view=agent" \
  -H "x-api-key: $SECAPI_API_KEY"
```

Section and semantic-search results can include `accession`, `section_key`, `source_url`, `char_start`, `char_end`, and `highlighted_snippet`. The character offsets refer to the returned section markdown, not raw SEC HTML. When offsets are unavailable, keep the accession and source URL but do not try to slice text; the response explains the degraded citation state.

## Use the result safely

1. Open the `source_url` or retrieve the section by `(accession, section_key)`.
2. Keep the accession number, filing date, and request ID beside your excerpt or model output.
3. Treat a match as evidence to review, not proof that a company has the risk, event, or accounting treatment you searched for.
4. Constrain repeatable work with ticker, form, date, or filing-year filters before expanding the corpus.

## Learn the exact contract

* [Full-text search reference](/api-reference/search/get-v1-search-fulltext)
* [Semantic search reference](/api-reference/search/get-v1-search-semantic)
* [Section search reference](/api-reference/sections/get-v1-sections-search)
* [Citation and response conventions](/api-conventions#citation-char-range-spans)
* [Semantic-search risk-factor tutorial](/tutorials/semantic-search-risk-factors)

For commercial context, see the [search feature page](https://secapi.ai/features/search), [semantic-search glossary entry](https://secapi.ai/glossary/semantic-search), and [SEC filing RAG guide](https://secapi.ai/guides/sec-filing-rag).